A carton production workshop order production management device

Abstract

The present invention discloses a carton production workshop order production management device, which is composed of a setting module, an acquisition module, a storage module, a modeling module, a calculation module and an output module. The basic parameter file is set by the setting module and stored in the storage module; the order information is received by the acquisition module and stored in the storage module; after receiving the generation instruction, the order data and the basic parameter file are read by the modeling module, and a mathematical model is formed according to the corresponding algorithm; in the calculation module, the optimization algorithm is called to obtain the production scheduling result, and it is forwarded to each production machine and production management platform through the output module. Through this application, the problems of low efficiency and incomplete consideration of order management in carton production workshops are solved, and the accuracy and efficiency of order management processing are achieved.

Description

Technical Field

[0001] The invention relates to an order management device for a carton production workshop and belongs to the field of carton production management. Background Art

[0002] In the carton production process, there are many varieties of cartons, small batches, many processes involved, low automation, and insufficient connection between processes. During the transfer of orders between processes, managers need to reallocate orders, which greatly reduces production efficiency. In this process, it is difficult for managers to consider the global optimality from the perspective of the overall production process, and the production efficiency is relatively low, which leads to a large room for optimization in production order scheduling.

[0003] With the continuous expansion of online shopping, the demand for cartons is increasing, and the orders are increasing, especially in small batches and multiple varieties. This puts forward new requirements for production management, not only to update production equipment and speed up production, but also to make higher requirements for the reasonable planning of order production. Summary of the invention

[0004] In order to overcome the shortcomings of the current order production management in carton production workshops, the present invention provides an order production management device for a carton production workshop. Under the constraints of meeting the production capacity of production machines and order production requirements, the optimal production goal of minimizing the overall production time of batch orders is achieved, thereby improving the overall production efficiency and reducing production costs. The present invention provides an order production management device for a carton production workshop.

[0005] The technical solution adopted by the present invention to solve its technical problems is: a carton production workshop order production management device, after receiving the order information, the device is modeled according to the pre-set specific conditions of each process equipment and the process requirements, quantity and delivery date of the order, and calculates the optimal order sorting that minimizes the maximum production time, and sends the results to each production machine for production; it is mainly composed of a setting module, an acquisition module, a storage module, a modeling module, a calculation module and an output module. Among them,

[0006] Setting module: Set basic parameters, including adding or reducing production machines and setting production machine parameters. When the workshop increases or decreases production machines, add or delete the type of production machine, the process to which it belongs, the number, the corresponding speed of the carton model, and other information.

[0007] Acquisition module: Receive order information, corresponding order raw material inventory information and real-time production status information of each production machine in the workshop.

[0008] Modeling module: Build an order scheduling model based on the information obtained by the acquisition module and the information set by the setting module.

[0009] Computing module: Apply hybrid particle swarm algorithm to solve the order scheduling model.

[0010] Storage module: stores the information set by the setting module, the information obtained by the acquisition module, and the information solved by the calculation module.

[0011] Output module: extracts the information solved by the calculation module to form management information and production machine information, and sends them to the management department and the corresponding production machine respectively.

[0012] Further, in the modeling module, the order scheduling model is constructed according to the information acquired by the acquisition module and the information set by the setting module by the following steps:

[0013] (2.1) According to the order information, determine the orders that can be scheduled with sufficient raw materials, determine the orders that need to be scheduled according to actual production needs, and renumber them {J1, J2, …, J n}, J is the order ID, n is the number of orders scheduled for production at that time;

[0014] (2.2) Extract the order process and the production machines involved in the process according to the process flow in the order. And according to the size of the product in the order and the order quantity, extract the speed and preparation time of the corresponding production machine, and calculate the time t required for the corresponding order to be produced by the corresponding production machine i,j,k , t i,j,k The time taken to process the i-th order on the k-th machine in the j-th process;

[0015] (2.3) Extract the information of the order production machine according to the process flow to form the machine allocation vector {M 1,1 ,M 1,2 ,...,M i,j} represents the production machine number of the jth process of the i-th order. Generate the process vector {O 1,1 ,O 1,2 ,...,O i,j} represents the jth process of the i-th order;

[0016] (2.4) According to the basic parameters set by the setting module and the calculation information of steps (2.1) and (2.2), the constraints and objective functions of the model are generated to extract the data required by the model. The required model is obtained. The description of the mathematical model is as follows:

[0017]

[0018] st S i,1 ≥0, i=1,2,...,n

[0019]

[0020] C i,j ≤S i,j+1 , i=1,2,...,n; j=1,2,...,n i -1

[0021]

[0022]

[0023]

[0024]

[0025]

[0026]

[0027]

[0028] The symbols are explained as follows:

[0029] i represents the order quantity;

[0030] j represents the process number;

[0031] k represents the number of production machines;

[0032] r represents the corresponding number of processes on a production machine;

[0033] S i,j Representative process O i,j Processing start time;

[0034] C i,j Representative process O i,j Processing end time;

[0035] x i,j,k Representative Order J i In O i,j Process M k Processing on machines;

[0036] y i,j,k,r Representative Order J i In O i,j Process M k The rth process on the machine;

[0037] Represents machine M k The start time of the previous r-th process;

[0038] q k Represents machine M kThe number of processing steps;

[0039] L is a positive constant.

[0040] The first constraint is that for any order J i , the processing preparation time starts from zero, and the starting time is greater than or equal to zero. Some carton production equipment needs to be debugged before production, and processing can start immediately after debugging. The second constraint formula shows the relationship between the start time and the end time of the production process. The third constraint formula shows that only the corresponding process O i,j After completion, the next process is O i,j+1 The fourth and fifth constraints indicate that for any process O i,j , it can only be processed by one machine, which requires that the number of carton orders cannot be set too large. The sixth constraint shows that there is only one machine M k After the rth process on the upper layer is completed, the r+1th process on it will start processing. The seventh and eighth constraints show that S i,j and The relationship between i,j It's Machine M k When the rth processing step is performed,

[0041] Further, in the calculation module, the application of the hybrid particle swarm algorithm to solve the order scheduling model is achieved by the following steps:

[0042] (3.1) Initialize the algorithm parameters. A particle i is represented by the process sequence vector and the machine allocation vector. Initialize the process sequence vector and the machine allocation vector. The particle swarm size is N1. The maximum iteration parameter required by the PSO algorithm is T1. The inertia weight w, the corresponding acceleration constants c1 and c2, the parameter for judging stagnation E, the maximum number of iterations of the taboo search algorithm is T2, the candidate set size is N2, and the taboo length is L.

[0043] (3.2) Set the iteration count t1 of the PSO algorithm to zero 0, the stagnation e to 0, randomly initialize the machine allocation vector of N1 particles according to the setting parameters and process flow, and obtain the process sequence vector according to step (3.5);

[0044] (3.3) Calculate the fitness of each particle i i , i = 1, 2, ..., N, the fitness is the production scheduling plan corresponding to the particle, that is, the completion time corresponding to the process sequence vector and the machine allocation vector Fitness fit = C, sj i is the completion time of the i-th order;

[0045] (3.4) Initialize pbest and gbest: For each particle i, let pbest i =fit i ,i=1,2,…,N. make pbest is the particle swarm optimal solution, gbest is the global optimal solution;

[0046] (3.5) Let t = t + 1, according to The particle is moved by the formula; where c1, c2 represent acceleration factors, r1, r2 are random numbers between 0 and 1, t is the current iteration number, and the operation express Cross with probability a and vector b, p i 0 and p g 0 Assign vectors to two randomly initialized machines that meet the production process. That is, by setting a decision vector of the same length as the machine allocation vector, the vector element value is 0 or 1, and the original machine allocation vector is cross-operated with a certain probability, and then the order production order of the corresponding production machine is arranged according to the principle of priority for the earlier process and the longer operation time, and then the process sequence vector is obtained;

[0047] (3.6) Evaluate the fitness of particle i i ,i=1,2,…,N;

[0048] (3.7) Update pbest: For each particle i, for fit i and pbest i For comparison, if fit i >pbest i , let pbest i =fit i , otherwise keep pbest i constant;

[0049] (3.8) Update gbest: If Just let Let e ​​= 0, otherwise let e = e + 1;

[0050] (3.9). If e ≥ E, perform the taboo search algorithm, use the particle corresponding to gbest as the initial solution of the taboo search algorithm, replace the particle with the search result, update the pbest corresponding to the particle, and update gbest, let e = 0, randomly select 70% of the particles from the remaining N1-1 particles, and then randomly initialize. Otherwise, go to step (3.10);

[0051] (3.10) If t<T1, return to step (3.5), otherwise end.

[0052] Further, the step (3.3) comprises the following steps:

[0053] (3.3.1) Initialization: Assume that the operable time of each order is Assume that the available time of each production machine is Let index = 0;

[0054] (3.3.2) Determine the current processing step: Let index = index + 1, and take the index-th element O in the process sequence vector. i,j For the currently processed process, determine the current production machine M according to the allocation vector of the production machine k ;

[0055] (3.3.3) Determine process O i,j Start time and completion time of process O i,j It can only be performed after the previous process of the same order and the previous production machine of the same order are completed. i,j The start time and end time are S i,j =max(sj i ,sm k ), C i,j =S i,j +t i,j,k ;

[0056] (3.3.4) Update sj i and sm k :Confirm O i,j Start and end time of order J i The operational time and machine M k The available time, let sj i =C i,j , sm k =C i,j ;

[0057] (3.3.5) If Then jump to step (3.3.2), otherwise calculate the completion time of the corresponding production scheduling plan order Let the fitness value fit = C.

[0058] Furthermore, the step (3.5) includes the following sub-steps:

[0059] (3.5.1) The production sequence of the production machines shall be arranged according to the principle of priority given to the smaller process J and priority given to the longer production time;

[0060] (3.5.2) Initialize two vectors pv and mv to identify the process and production sequence of the order, respectively. The length is equal to the process sequence vector. In the initial state, the vector pv represents O i,1 The position where (i=1,2,...,n) is located is 1, and the rest are 0. mv is the position where the first order executed on each production machine is 1, and the rest are 0;

[0061] (3.5.3) Randomly select a process O whose pv and mv are both 1 i,j , there is a machine allocation vector that determines the production machine M that executes it k ;

[0062] (3.5.4) Update process O i,j The start time, end time, and machine M k The time set of the order execution time and the execution time of the order; update the pv vector, let process O i,j+1 The corresponding element is 1, M k Production machine process O i,j The position of the next process element is set to 1:

[0063] (3.5.5) Repeat (3.5.3)-(3.5.4) until all operations are calculated: finally determine the process sequence vector based on the start time.

[0064] The beneficial effects of the present invention are that the present invention sets the basic parameter file through the setting module and stores it in the storage module; receives the order information through the acquisition module and stores it in the storage module; after receiving the generation instruction, the order data and the basic parameter file are read through the modeling module, and a mathematical model is formed according to the corresponding algorithm; in the calculation module, the optimization algorithm is called to obtain the production scheduling result, and it is forwarded to each production machine and production management platform through the output module. Through this application, the problems of low efficiency and incomplete consideration of order management in the carton production workshop are solved, and the accuracy and efficiency of order management processing are achieved. BRIEF DESCRIPTION OF THE DRAWINGS

[0065] Figure 1 It is a structural diagram of the device of the present invention;

[0066] Figure 2 It is an algorithm flow chart of the calculation module of the present invention. DETAILED DESCRIPTION

[0067] The following is based on the attached Figure 1 The device of the present invention is specifically described as an order production management device for a carton production workshop.

[0068] The present invention is further described in detail below with reference to examples, but the implementation manner of the present invention is not limited to these examples.

[0069] The implementation of the present invention is described in detail below in conjunction with specific embodiments.

[0070] The carton production workshop is a flexible assembly line workshop. The carton production process includes corrugated carton production line, mounting, die-cutting, stapling, gluing and packaging. Stapling and gluing are the same process. The same machine can only process one process at a time, and the order of processing between orders is not fixed. Each order determines the production process according to its corresponding process requirements.

[0071] like Figure 1 The present invention shows a carton production workshop order production management device, and the specific modules are as follows:

[0072] Setting module 1, setting basic parameters, including adding or reducing production machines and setting production machine parameters. When the workshop adds or reduces production machines, add or delete the type of production machine, the process to which it belongs, the number, the corresponding speed of the carton model, and other information;

[0073] Acquisition module 2 receives order information, inventory information of corresponding order raw materials and real-time information on the production status of each production machine in the workshop;

[0074] The order information is listed below:

[0075] Order Number Process area quantity

[0076] Modeling module 3, builds an order scheduling model based on order information and data set by the setting module;

[0077] Computation module 4, using hybrid particle swarm algorithm to solve the order scheduling model;

[0078] Storage module 5, storing the results obtained by the setting module, the acquisition module and the calculation module into the storage module for use and query;

[0079] The output module 6 extracts the calculation results to form overall information and production machine information, and sends them to the management department and the corresponding production machine respectively.

[0080] Further, for the setting module 1, the setting contents are as follows:

[0081] Machine set: {M1, M2, …, M m}, corresponding to the process to which it belongs, and the speed of the corresponding machine at the corresponding carton size.

[0082] Set the corresponding initial time for each machine based on experience. For example:

[0083]

[0084] Process set: {Oi,1 , O i,2 ,…,O i,n}, indicating the process of the i-th order, the order of the same process, the process has been determined.

[0085] For example:

[0086]

[0087] The modeling module 3 includes production machine determination processing, production machine information extraction processing, constraint condition determination, and objective function determination. Its implementation steps include:

[0088] (2.1) According to the order information, determine the orders that can be scheduled with sufficient raw materials, determine the orders that need to be scheduled according to actual production needs, and renumber them {J1, J2, …, J n}, J is the order ID, n is the number of orders scheduled for production at that time;

[0089] (2.2) Extract the order process and the production machines involved in the process according to the process flow in the order. And according to the size of the product in the order and the order quantity, extract the speed and preparation time of the corresponding production machine, and calculate the time t required for the corresponding order to be produced by the corresponding production machine i,j,k , t i,j,k The time taken to process the i-th order on the k-th machine in the j-th process;

[0090] (2.3) Extract the information of the order production machine according to the process flow to form the machine allocation vector {M 1,1 ,M 1,2 ,...,M i,j} represents the production machine number of the jth process of the i-th order. Generate the process vector {O 1,1 ,O 1,2 ,...,O i,j} represents the jth process of the i-th order;

[0091] (2.4) Based on the preset parameters and the calculation information of (2.1) and (2.2), the constraints and objective functions of the model are generated to extract the data required by the model. The required model is obtained. The description of the mathematical model is as follows:

[0092]

[0093] st S i,1 ≥0, i=1,2,...,n

[0094]

[0095] C i,j ≤S i,j+1, i=1,2,...,n; j=1,2,...,n i -1

[0096]

[0097]

[0098]

[0099]

[0100]

[0101]

[0102]

[0103] The symbols are explained as follows:

[0104] i represents the order quantity;

[0105] j represents the process number;

[0106] k represents the number of production machines;

[0107] r represents the corresponding number of processes on a production machine;

[0108] S i,j Representative process O i,j Processing start time;

[0109] C i,j Representative process O i,j Processing end time;

[0110] x i,j,k Representative Order J i In O i,j Process M k Processing on machines;

[0111] y i,j,k,r Representative Order J i In O i,j Process M k The rth process on the machine;

[0112] Indicates machine M k The start time of the previous r-th process;

[0113] q k Indicates machine M k The number of processing steps;

[0114] L is a positive constant.

[0115] The first constraint is that for any order J i , the processing preparation time starts from zero, and the starting time is greater than or equal to zero. Some carton production equipment needs to be debugged before production, and processing can start immediately after debugging. The second constraint formula shows the relationship between the start time and the end time of the production process. The third constraint formula shows that only the corresponding process O i,j After completion, the next process is O i,j+1 The fourth and fifth constraints indicate that for any process O i,j , it can only be processed by one machine, which requires that the number of carton orders cannot be set too large. The sixth constraint shows that there is only one machine M k After the rth process on the upper layer is completed, the r+1th process on it will start processing. The seventh and eighth constraints show that S i,j and The relationship between i,j It's Machine M k When the rth processing step is performed,

[0116] The calculation module 4 is characterized by improving the particle swarm algorithm to realize the hybrid particle swarm algorithm; at the same time, a taboo search link is added to the algorithm to avoid local optimality and improve the efficiency of the algorithm, thereby finally realizing the hybrid particle swarm algorithm. Figure 2 As shown, the steps are as follows:

[0117] (3.1) Initialize the algorithm parameters. A particle i is represented by the process sequence vector and the machine allocation vector. Initialize the process sequence vector and the machine allocation vector. The particle swarm size is N1. The maximum iteration parameter required by the PSO algorithm is T1. The inertia weight w, the corresponding acceleration constants c1 and c2, the parameter for judging stagnation E, the maximum number of iterations of the taboo search algorithm is T2, the candidate set size is N2, and the taboo length is L.

[0118] (3.2) Set the iteration count t1 of the PSO algorithm to zero 0, the stagnation e to 0, randomly initialize the machine allocation vector of N1 particles according to the setting parameters and process flow, and obtain the process sequence vector according to step (3.5);

[0119] (3.3) Calculate the fitness of each particle i i , i = 1, 2, ..., N, the fitness is the production scheduling plan corresponding to the particle, that is, the completion time corresponding to the process sequence vector and the machine allocation vector Fitness fit = C, sj i is the completion time of the i-th order;

[0120] (3.4) Initialize pbest and gbest: For each particle i, let pbest i =fit i ,i=1,2,…,N. make pbest is the particle swarm optimal solution, gbest is the global optimal solution;

[0121] (3.5) Let t = t + 1, according to The particle is moved by the formula; where c1, c2 represent acceleration factors, r1, r2 are random numbers between 0 and 1, t is the current iteration number, and the operation express Cross with probability a and vector b, p i 0 and p g 0 Assign vectors to two randomly initialized machines that meet the production process. That is, by setting a decision vector of the same length as the machine allocation vector, the vector element value is 0 or 1, and the original machine allocation vector is cross-operated with a certain probability, and then the order production order of the corresponding production machine is arranged according to the principle of priority for the earlier process and the longer operation time, and then the process sequence vector is obtained;

[0122] (3.6) Evaluate the fitness of particle i i ,i=1,2,…,N;

[0123] (3.7) Update pbest: For each particle i, for fit i and pbest i For comparison, if fit i >pbest i , let pbest i =fit i , otherwise keep pbest i constant;

[0124] (3.8) Update gbest: If Just let Let e ​​= 0, otherwise let e = e + 1;

[0125] (3.9). If e ≥ E, perform the taboo search algorithm, use the particle corresponding to gbest as the initial solution of the taboo search algorithm, replace the particle with the search result, update the pbest corresponding to the particle, and update gbest, let e = 0, randomly select 70% of the particles from the remaining N1-1 particles, and then randomly initialize. Otherwise, go to step (3.10);

[0126] (3.10) If t<T1, return to step (3.5), otherwise end.

[0127] Regarding step (3.3), the fitness fit is required. The specific steps are as follows:

[0128] (3.3.1) Initialization: Assume that the operable time of each order is Assume that the available time of each production machine is Let index = 0;

[0129] (3.3.2) Determine the current processing step: Let index = index + 1, and take the index-th element O in the process sequence vector. i,j For the currently processed process, determine the current production machine M according to the allocation vector of the production machine k ;

[0130] (3.3.3) Determine process O i,j Start time and completion time of process O i,j It can only be performed after the previous process of the same order and the previous production machine of the same order are completed. i,j The start time and end time are S i,j =max(sj i ,sm k ), C i,j =S i,j +t i,j,k ;

[0131] (3.3.4) Update sj i and sm k :Confirm O i,j Start and end time of order J i The operational time and machine M k The available time, let sj i =C i,j , sm k =C i,j ;

[0132] (3.3.5) If Then jump to step (3.3.2), otherwise calculate the completion time of the corresponding production scheduling plan order Let the fitness value fit = C.

[0133] In step (3.5), the process sequence vector is obtained from the machine allocation vector. The specific steps are as follows:

[0134] (3.5.1) The production sequence of the production machines shall be arranged according to the principle of priority given to the smaller process J and priority given to the longer production time;

[0135] (3.5.2) Initialize two vectors pv and mv to identify the process and production sequence of the order, respectively. The length is equal to the process sequence vector. In the initial state, the vector pv represents O i,1 The position where (i=1,2,...,n) is located is 1, and the rest are 0. mv is the position where the first order executed on each production machine is 1, and the rest are 0;

[0136] (3.5.3) Randomly select a process O whose pv and mv are both 1 i,j , there is a machine allocation vector that determines the production machine M that executes it k ;

[0137] (3.5.4) Update process O i,j The start time, end time, and machine M k The time set of the order execution time and the execution time of the order; update the pv vector, let process O i,j+1 The corresponding element is 1, M k Production machine process O i,j The element position of the next process is set to 1;

[0138] (3.5.5) Repeat steps (3.5.3)-(3.5.4) until all operations are calculated: finally, determine the process sequence vector based on the start time.

[0139] After the above steps are completed, the management department obtains the production order scheduling results, forms the corresponding Gantt chart, and each production machine obtains the corresponding production machine order production sequence.

[0140] The above embodiments are used to illustrate the present invention rather than to limit the present invention. Any modification and change made to the present invention within the spirit of the present invention and the protection scope of the claims shall fall within the protection scope of the present invention.

Claims

1. A carton production workshop order production management device, after receiving order information, the device models according to the pre-set specific conditions of each process equipment and the process requirements, quantity and delivery date data of the order, and calculates the optimal order sequence that minimizes the maximum production time, and sends the result to each production machine for production; it is characterized by: It consists of a setting module, an acquisition module, a storage module, a modeling module, a calculation module and an output module; among them, Setting module: set basic parameters, including adding or reducing production machines and setting production machine parameters; when the workshop increases or decreases production machines, add or delete the type, process, number, carton model and corresponding speed information of the production machine; Acquisition module: receiving order information, inventory information of corresponding order raw materials and real-time information on the production status of each production machine in the workshop; Modeling module: builds an order scheduling model based on the information obtained by the acquisition module and the information set by the setting module; Computing module: Apply hybrid particle swarm algorithm to solve the order scheduling model; Storage module: stores information set by the setting module, information obtained by the acquisition module, and information solved by the calculation module; Output module: extracts the information solved by the calculation module to form management information and production machine information, and sends them to the management department and the corresponding production machine respectively; In the modeling module, the order scheduling model is constructed according to the information acquired by the acquisition module and the information set by the setting module, which is achieved by the following steps: (2.1) According to the order information, determine the orders that can be scheduled with sufficient raw materials, determine the orders that need to be scheduled according to actual production needs, and renumber them {J1, J2, …, J i ,…J n }, J is the order ID, n is the number of orders scheduled for production at that time; (2.2) Extract the order process and the production machines involved in the process according to the process flow in the order; and extract the speed and preparation time of the corresponding production machine according to the size of the product and the order quantity in the order, and calculate the time t required for the corresponding order to be produced by the corresponding production machine i,j,k , t i,j,k The time taken to process the i-th order on the k-th machine in the j-th process; (2.3) Extract the information of the order production machine according to the process flow to form the machine allocation vector {M 1,1 ,M 1,2 ,...,M i,j } represents the production machine number of the jth process of the i-th order; Generate the process vector of all production orders {O 1,1 ,O 1,2 ,...,O i,j } represents the jth process of the i-th order; (2.4) Based on the basic parameters set by the setting module and the calculation information of steps (2.1) and (2.2), the constraints and objective functions of the model are generated to extract the data required by the model; the required model is obtained. The description of the mathematical model is as follows: The symbols are explained as follows: i represents the order quantity; j represents the process number; k represents the number of production machines; r represents the corresponding number of processes on a production machine; S i,j Representative process O i,j Processing start time; C i,j Representative process O i,j Processing end time; x i,j,k Representative Order J i In O i,j Process M k Processing on machines; y i,j,k,r Representative Order J i In O i,j Process M k The rth process on the machine; Represents machine M k The start time of the previous r-th process; q k Represents machine M k The number of processing steps; L is the taboo length, which is a positive constant; The first constraint is that for any order J i , the processing preparation time starts from time zero, and the starting time is greater than or equal to zero. Some carton production equipment needs to be debugged before production, and processing can start immediately after debugging; the second constraint formula shows the relationship between the start time and the end time of the production process; the third constraint formula shows that only the corresponding process O i,j After completion, the next process is O i,j+1 The fourth and fifth constraints indicate that for any process O i,j , it can only be processed by one machine; the sixth constraint shows that there is only one machine M k After the rth process on the upper layer is completed, the r+1th process on it will start processing; the seventh and eighth constraint equations show that S i,j and The relationship between i,j It's Machine M k When the rth processing step is performed, In the calculation module, the application of the hybrid particle swarm algorithm to solve the order scheduling model is achieved by the following steps: (3.1) Initialize the algorithm parameters. A particle σ is represented by the process sequence vector and the machine allocation vector. Initialize the process sequence vector and the machine allocation vector. The particle swarm size is N1. The maximum iteration parameter required by the PSO algorithm is T1. The inertia weight w, the corresponding acceleration constants c1 and c2, the parameter for judging stagnation E, the maximum number of iterations of the taboo search algorithm is T2, the candidate set size is N2, and the taboo length is L. (3.2) Set the iteration count t1 of the PSO algorithm to zero 0, the stagnation e to 0, randomly initialize the machine allocation vector of N1 particles according to the setting parameters and process flow, and obtain the process sequence vector according to step (3.5); (3.3) Calculate the fitness of each particle σfit σ ,σ=1,2,…,N, the fitness is the production scheduling plan corresponding to the particle, that is, the completion time corresponding to the process sequence vector and the machine allocation vector Fitness fit = C, sj i is the operable time of the i-th order; (3.4) Initialize pbest and gbest: For each particle σ, let pbest σ =fit σ ,σ=1,2,…,N; let pbest is the particle swarm optimal solution, gbest is the global optimal solution; (3.5) Let t = t + 1, according to The particle is moved by the formula; where c1, c2 represent acceleration factors, r1, r2 are random numbers between 0 and 1, t is the current iteration number, and the operation express With probability c1r1 and vector p i t To cross, express With probability c2r2 and vector p g t Crossover, p i 0 and p g 0 Assign vectors to two randomly initialized machines that meet the production process; that is, by setting a decision vector with the same length as the machine assignment vector, the vector element value is 0 or 1, and the original machine assignment vector is cross-operated with a certain probability, and then the order production order of the corresponding production machines is arranged according to the principle of priority for the earlier process and the longer operation time, and then the process sequence vector is obtained; (3.6) Evaluate the fitness of particle σfit σ ,σ=1,2,…,N; (3.7) Update pbest: For each particle σ, for fit σ and pbest σ For comparison, if fit σ >pbest σ , let pbest σ =fit σ , otherwise keep pbest σ constant; (3.8) Update gbest: If Just let Let e ​​= 0, otherwise let e = e + 1; (3.9) If e ≥ E, perform the taboo search algorithm, use the particle corresponding to gbest as the initial solution of the taboo search algorithm, replace the particle with the search result, update the pbest corresponding to the particle, and update gbest, let e = 0, randomly select 70% of the particles from the remaining N1-1 particles, and then randomly initialize; otherwise, go to step (3.10); (3.10) If t<T1, return to step (3.5), otherwise end.

2. According to the carton production workshop order production management device of claim 1, the step (3.3) comprises the following sub-steps: (3.3.1) Initialization: Let the operable time of each order be sj i =0, Assume that the available time of each production machine is sm k =0, Let j = 0; (3.3.2) Determine the current processing step: Let j = j + 1, take the jth element O in the process sequence vector i,j For the currently processed process, determine the current production machine M according to the allocation vector of the production machine k ; (3.3.3) Determine process O i,j Start time and completion time of process O i,j It can only be executed after the previous process of the same order and the previous production machine of the same order are completed; therefore, the operation O i,j The start time and end time are S i,j =max(sj i ,sm k ), C i,j =S i,j +t i,j,k ; (3.3.4) Update sj i and sm k :Confirm O i,j start and end time, update any order J i The operational time and machine M k The available time, let sj i =C i,j , sm k =C i,j ; (3.3.5) If Then jump to step (3.3.2), otherwise calculate the completion time corresponding to the process sequence vector and the machine allocation vector Let the fitness value fit = C.

3. According to the carton production workshop order production management device of claim 2, the step (3.5) comprises the following sub-steps: (3.5.1) The production sequence of the production machines is arranged according to the principle of priority for smaller process j and priority for longer production time; (3.5.2) Initialize two vectors pv and mv to identify the process and production sequence of the order, respectively. The length is equal to the process sequence vector. In the initial state, the vector pv represents O i,1 , i=1,2,...,n, the position is 1, and the rest are 0; (3.5.3) Randomly select a process O whose pv and mv are both 1 i,j , there is a machine allocation vector that determines the production machine M that executes it k ; (3.5.4) Update process O i,j The start time, end time, and machine M k The time set of the order execution time and the execution time of the order; update the pv vector, let process O i,j+1 The corresponding element is 1, M k Production machine process O i,j The position of the next process element is set to 1: (3.5.5) Repeat (3.5.3)-(3.5.4) until all operations are calculated: finally determine the process sequence vector based on the start time.

Images